The most difficult area of IT to deal with, is those systems which are neither novel and new (i.e. some examples of the service are already in existence) nor common and ubiquitous.
Now what is key here, is to know how you to got to this position.
If you created a CA like system which is now being copied by other competitors (presumably because of your success) you run the risk of the cost of migration.
This is where someone releases a product that becomes the standard in your industry and you are constantly maintaining and upgrading your service to keep up with the standard. You will almost certainly in the long run have to switch to the standard - so your costs will be the switch and the cost of continuing your own in-house product until you made the switch. This cost can be very significant.
The best tactical approach to secure the most value in such a situation, is to release your service as a product when other competitors start copying your service. Give it to someone to sell, give it away, open source it etc. If your service becomes the standard, then there is no cost of migration. Open source is potentially a powerful tactic in this context.
Now, if you are instead a laggard into this new field and you see that some of your competitors have such a service whilst no standard product exists then you have two choices. The disruption game or the wait and adopt approach.
Both you'll have to balance carefully in terms of worth, cost, risk and migration effects.
In the disruption game - you try to be the first to release a product and to get that product adopted as the standard. Steal your competitors teams, pay them to develop and release a modified version etc. It's a risky play and it all depends upon the service - whether it could become some form of strategically important platform, like a universal airline booking system? How much influence do you want to purchase over its future direction? etc
The wait and adopt, is just that. Don't do anything and wait until a standard product is released, then adopt it.
In this class of problem, the ROI syndrome is very much secondary to the pure tactical play.
